Analytics Experts Offer Advice on Building the Ultimate March Madness Bracket

There’s no escaping the buzz about sports and data. Almost every major sport is interested in collecting, parsing and understanding the way their game is played in the hope that hidden, game-changing truths can be discovered.

STATS, a sports and technology company, has been at the forefront of this trend in basketball thanks to its impressive SportVU solution, which offers in-depth player tracking technology that can help basketball coaches identify problems with a player’s gameplay — problems that might otherwise have gone unnoticed.

In celebration of March Madness, the NCAA Men’s Division I Basketball Championship, CDW sponsored a Q&A on Reddit with Ethan Cooperson, a senior research analyst from STATS, and retired NBA player and ESPN analyst Jay Williams. The two offered advice on how college basketball fans can use sports data to put together winning March Madness brackets.

One Reddit user asked if it was best to put out a “stupid bracket” — one that few people would consider winnable — that took a wild stab at an obvious bracket. Cooperson suggested that looking at wild-card players, rather than teams, was a better way to go about “predicting the unpredictable.”

“Perhaps there is a single player who can take over and carry his team deep into the Tourney. Consider Kemba Walker with UConn in 2011 (23.5 points/gm, 5.7 assists/gm); that was not a great team throughout the season, but Walker carried them to the title. Rather than trying to predict a slew of upsets, perhaps pick a single team — and player — with the capability to rise up in March/April,” he responded.

Another Reddit user asked if historical performance should factor into the decision-making that goes into putting together a bracket.

“Last 5 years, of the 20 Final Four teams, only 2 of them entered the Tourney with a win streak of 10+ games (2010 Butler, 2013 Louisville). Of those 20, 7 of them were no better than 6-4 in their 10 games prior to the Tourney. So it appears that last 10 games are not a great indicator. Last year, both Michigan and Syracuse were 5-5 in their last 10 entering the Tourney and reached the Final Four,” Cooperson explained.

This lines up with recent research that BizTech reported on that found that there’s no such thing as a hot streak in basketball.

“Unless a team is returning a major portion of their cast from the year prior I would not base my projections off of historical odds,” Williams added.

As far as which teams were poised to be “bracket busters” in this year’s tournament: “North Dakota St, Harvard, Pittsburgh, Oklahoma State, Iowa State, & Mercer could give Duke an issue,” Williams suggested.

While numbers, stats and data are exciting additions to the sports entertainment industry, for Williams, it doesn’t replace the gut instinct that comes with being a basketball fan: “Using metrics to evaluate the talent of individuals and teams are needed most definitely. I based a ton of my breakdowns via broadcasting with numbers but there is also a feel to the game or any sport. There has to be some sort of balance when evaluating.”